Download presentation
Presentation is loading. Please wait.
Published byAlexandra Bradley Modified over 11 years ago
1
Libra: An Economy driven Job Scheduling System for Clusters Jahanzeb Sherwani 1, Nosheen Ali 1, Nausheen Lotia 1, Zahra Hayat 1, Rajkumar Buyya 2 1. Lahore University of Science and Management (LUMS), Lahore, Pakistan 2. Grid Computing and Distributed Systems (GRIDS) Lab., University of Melbourne, Australia www.gridbus.org
2
2 Agenda Introduction/Motivations The Libra Scheduler Architecture & Cost- based Scheduling Strategy Implementation Performance Evaluation Conclusion and Future Work
3
3 Introduction Clusters (of commodity computers) have emerged as mainstream parallel and distributed platforms for high performance, high-throughput and high- availability computing. They have been used in solving numerous problems in science, engineering, and commerce.
4
4 Adoption of the Approach Oracle
5
5 Cluster Resource Management System: Managing the Shared Facility Sequential Applications Parallel Applications Parallel Programming Environment Cluster Management System (Single System Image and Availability Infrastructure) Cluster Interconnection Network/Switch PC/Workstation Network Interface Hardware Communications Software PC/Workstation Network Interface Hardware Communications Software PC/Workstation Network Interface Hardware Communications Software PC/Workstation Network Interface Hardware Communications Software Sequential Applications Parallel Applications
6
6 Some Cluster Management Systems Commercial and Open-source Cluster Management Software Open-source Cluster Management Software DQS (Distributed Queuing System ) Condor GNQS (Generalized Network Queuing System) MOSIX Load Leveler SGE (Sun Grid Engine) PBS (Portable Batch System)
7
7 Cluster Management Systems Still Use System Centric Approach Traditional CMSs focus has essentially been on maximizing CPU performance, but not on improving the value of utility delivered to the user and quality of services. Traditional system-centric performance metrics CPU Throughput Mean Response Time Shortest Job First FCFS Some Static Priorities …
8
8 The Libra Approach: Computational Economy Paradigm for Management & Job Scheduling
9
9 Cost Model: Why are they needed ? Without cost model any shared system becomes un-manageable It supports QoS based resource allocation and help manage supply-and-demand for resources. Improves the value of utility delivered. Also, improves the resource utilization. Cost units (G$) may be Rupees/Dollars (real money) Shares in global facility Stored in bank
10
10 Cost Matrix Non-uniform costing Different users are charged different prices that vary with time. 13 21 User 5 Machine 1 User 1 Machine 5 Resource Cost = Function (cpu, memory, disk, network, software, QoS, current demand, etc.) Simple: price based on peaktime, offpeak, discount when less demand,..
11
11 Computational Economy Parameters Job parameters most relevant to user- centric scheduling Budget allocated to job by user Deadline specified by user
12
12 Libra Architecture User Cluster Management System (PBS) Job Input Control Application Scheduler (Libra) Budget Check Control Deadline Control Server: Master Node Node Querying Module Best Node Evaluator Job Dispatch Control Cluster Worker Nodes with node monitor (pbs_mom) User Application (node 1) (node 2) (node N)... (job, deadline, budget)
13
13 Libra with PBS Portable Batch System (PBS) as the Cluster Management Software (CMS) Robust, portable, effective, extensible batch job queuing and resource management system Supports different schedulers Job accounting Allows Plugging of Third-Party Scheduling Solution
14
14 The Libra Scheduler Job Input Controller Adding parameters at job submission time deadline budget Execution Time Defining new attributes of job Job Acceptance and Assignment Controller Budget checked through cost function Admission control through deadline scheduling Execution host with the minimum load and ability to finish job on time selected Node Resource Share Allocation: Proportional to the needs of multiple User Jobs QoS needs.
15
15 The Libra Scheduler Job Execution Controller Job run on the best node according to algorithm Cluster and node status updated runTime cpuLoad Job Querying Controller Server, Scheduler, Exec Host, and Accounting Logs
16
16 Pricing the Cluster Resources Cost= α* (Job Execution Time) + β* (Job Execution Time / Deadline) Cost = α*E + β*E/D (where α and β are coefficients) Cost of using the cluster depends on job length and job deadline: the longer the user is prepared to wait for the results, the lower his cost Cost formula motivates users to reveal their true QoS requirements (e.g., deadline)
17
17 PBS-Libra Web --- Front-end for the Libra Engine
18
18 PBS- Libra Web
19
19 PBS- Libra Web
20
20 Performance Evaluation: Simulations Goal: Measure the performance of Libra Scheduler Performance = ? Maximize user satisfaction Maximise value delivered by the utility Simulation Platform: GridSim Simulated scheduling using the GridSim toolkit http://www.gridbus.org/gridsim
21
21 Simulations Methodology Workload 120 jobs with deadlines and budgets Job lengths: 1000 to 10000 (MIs) Resources 10 node, single processor (MIPS rating: 100) (homogenous) cluster
22
22 Simulations Scheduler simulated as a function Input: job size, deadline, budget Output: accept/reject, node #, share allocated
23
23 Simulations Compared: Proportional Share (Libra) FIFO (PBS) Experiments: 120 jobs, 10 nodes Increasing workload to 150 and 200 Increasing cluster size to 20
24
24 Simulation Results 120 jobs, 20 did not meet budget
25
25 100 Jobs, 10 Nodes FIFO: 23 rejected - Proportional Share: 14 rejected Deadline Completion time. PBS FIFO Libra Proportional
26
26 Simulation Results Increase workload to 200 jobs on the same 10 node cluster
27
27 200 Jobs, 10 Nodes FIFO: 105 rejected - Proportional Share: 93 rejected PBS FIFO Libra Proportional
28
28 Simulation Results Scale the cluster up to 20 nodes
29
29 200 Jobs, 20 Nodes FIFO: 35 rejected - Proportional Share: 23 rejected
30
30 PBS FIFO & Libra Strategy No. of Jobs No. of Nodes No. of Jobs AcceptedNo. of Jobs Rejected PBS FIFOLibraPBS FIFOLibra 100 1077862314 2090861014 200 109510210598 201651773523
31
31 Conclusion & Future Work Successfully developed a Linux-based cluster that schedules jobs using PBS with our economy- driven Libra scheduler, and PBS-Libra Web as the front end. Successfully tested our scheduling policy Proportional Share delivers more value to users Exploring other pricing mechanisms Expanding the cluster with more nodes and with support for parallel jobs Implement Libra for SGE (Sun Grid Engine) Sponsored by Sun!
32
32 Thank you
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.